Current camera based systems used to track or trace the trajectories of sports objects suffer from all the known shortcomings of image processing. For example a golf ball may be difficult to detect or measure in a camera image due to contrast, ambient light conditions, and camera resolution. Current systems that provide traces of ball flight on television feeds or tennis ball trajectories for line calling decisions use image processing techniques to find the ball in successive images and to construct a trace or trajectory from such finds. Apart from being able to animate the television images with traces of the ball flight, accurate data such as positions, speeds or spin rates are not readily available from such systems.
Current radar based systems used to track sports objects suffer from other shortcomings unique to radar technology. For example the vertical angular position or height of an object at low elevation can be severely affected by multipath effects.
There is a need to improve both camera and radar tracking systems for sports objects.